Ryan Marten - Vertical Federated Learning

Показать описание

Vector interns are curious students and researchers affiliated with the Vector Institute who collaborate with the best and brightest AI developers, scaling research breakthroughs in Machine Learning and Deep Learning.

Below is one of many Internship Talks, where interns presented their projects after completing their work term. Projects span across sectors, some of which include: Project Management, Automation, Applied Machine Learning, and Data Visualization.

PETs: Privacy Enhancing Techniques, such as federated learning, differential privacy, and homomorphic encryption. Ryan’s talk focuses on Vertical Federated Learning as a Privacy Enhancing Technique.

There is a large amount of data stored by individual organizations which, if used collectively to train machine learning models, could benefit consumers or the general public. For example, if a model can be developed using patient data from all hospitals across Canada, we could understand the health situation of Canadians and allocate resources and funding both in research and in hospitals accordingly. It is evident that such models could help hospitals across Canada operate more efficiently and advance research where it is most needed. However, organizations such as hospitals are often unable to share data due to privacy constraints. As a result, how can we train a machine learning model leveraging data stored across organizations without creating privacy risks? In this work, we showcase how privacy preserving machine learning technologies (e.g. differential privacy, federated learning) enable the training of such models.

Vector Institute

Рекомендации по теме

Комментарии

hi how to choose from HE, DP, SMPC for Federated learning for privacy protection? Any standards to follow?

macknightxu

Ryan Marten - Vertical Federated Learning

Ryan Marten - Vertical Federated Learning

Vertical Federated Learning

Federated Learning: Privacy's Frontier

Vertical FL using Financial Data (Flower Monthly 2023-04)

ADI: Adversarial Dominating Inputs in Vertical Federated Learning Systems

USENIX Security '22 - Label Inference Attacks Against Vertical Federated Learning

FLOW Seminar #74: Bin Gu (MBZUAI) Fast and Secure Vertical Federated Learning

Federated Learning Strategies | Taxonomy

Horizontal Federated Learning

[9B] Comprehensive Analysis of Privacy Leakage in Vertical Federated Learning During Prediction

Vertical Federated Learning across the Value Chain | #BAS22

Privacy Preserving Asynchronous Vertical Federated Learning Algorithms for Multiparty Collaborative

Federated Learning

Introduction to the Workshop on Parallel, Distributed, and Federated Learning (Yamuna Krishnamurthy)

Best Paper Award at the workshop on Parallel, Distributed, and Federated Learning (PDFL) at ECMLPKDD

[Paper review] Incentive mechanism for reliable federated learning

What, Why and How of Federated Machine Learning – Implementation Using... Neeraj Arora & Layne ...

Federated Self-Supervised Speech Representation Learning (Flower Monthly 2023-09)

SHARE: Shaping Data Distribution at Edge for Communication-Efficient Hierarchical Federated Learning

Split learning for vertically partitioned data

OM PriCon2020 Tutorial: Privacy Preserving Vertically Distributed Machine Learning Tutorial

Privacy and Federated Learning with doc.ai

AIED 2021 Workshop: Qiang Yang 'Federated Machine Learning and XAI'

A demonstration of federated learning in Scaleout Studio